Pro-social rule breaking (PSRB), as one form of constructive deviance, is receiving increasing attention from researchers, yet surprisingly our knowledge of the antecedents of PSRB is fairly limited. Although the effects of personal traits on PSRB are well discussed in the literature, we know little about how contextual factors in the work environment, such as leadership style, shape PSRB. The current study aims to address this research gap by examining the relationship between ethical leadership and PSRB. Specifically, drawing from the social information processing theory, we propose a dual process model that ethical leadership enhances employees’ PSRB via psychological safety and job autonomy. We further investigate the moderating effect of employees’ risk taking propensity in the model. The participants were recruited from four organizations in three cities of China (i.e., Beijing, Shanghai, and Dalian). The data was collected at two time points to avoid the common method bias. At Time 1, ethical leadership, psychological safety, job autonomy, and risk taking propensity were measured. At Time 2, that is, one month later, participants’ PSRB behavior was assessed. The final sample size was 241. We used SPSS 22.0 and Mplus 6.12 to analyse the data. First, we assessed the discriminant validity of the key variables through a series of confirmatory factor analyses (CFAs) and examined the common method variance. Afterwards, we used the Structural Equation Modeling (SEM) and bias-corrected bootstrap to test our hypotheses. The results showed a significantly positive relationship between ethical leadership and PSRB, and the relationship was mediated by psychological safety and job autonomy. Furthermore, individual risk taking propensity moderated the relationship between job autonomy and PSRB, such that the positive relationship was stronger for those employees with higher, rather than lower, levels of risk taking propensity. Our study contributes to the PSRB literature in several ways. Frist, we investigate the effects of ethical leadership on employees’ PSRB, which broadens our understanding of how contextual factors influences PSRB. Second, we reveal the theoretical black-box of why ethical leadership facilitates PSRB by identifying psychological safety and job autonomy as two underlying mechanisms. Finally, we extend the ethical leadership and PSRB research by demonstrating risk taking propensity as a crucial contingency that moderates the effects of ethical leadership and job autonomy on PSRB.